Part of the reason that Toyota did not address the accelerator problem more decisively was the fact that the external sensors (data collection points) that they were utilizing were not well integrated and collected a lot of erroneous or misleading data. NHTSA was perhaps one of the biggest contributors to this problem by not feeding all accelerator issues back to Toyota, or providing collateral data indicating the problem may have been driver error and thus adding ‘noise’ to the data needed to fully understand the extent and the root of the problem.
Now in no way am I saying Toyota was innocent in this matter, but I do want to defend them in their methodic ways of not jumping to conclusions without letting the data speak first. One of the core teachings of Lean is that “you cannot improve what you cannot measure” however measuring also has to be accurate. If data is inaccurate, as was the case with the acceleration problem, then it is understandable that there was a delay before Toyota clearly saw that there was a major problem to be addressed.
The dilemma stems from the fact that this dirty data cannot be used as a defense. To a degree most auto manufacturers, and to that extent, most organizations are using sensors that are potentially flawed and inaccurate, let alone internal sensors. When it comes to safety matters there really is no room for error since bad data is as good as no data, and people’s lives are worth the extra time and effort to ensure data is accurate and the handling of the data is treated with priority.
Recently I had the opportunity to speak with someone at Schlumberger regarding their QUEST program. The quest program is an idea management system focused on safety issues and based on the DuPont STOP methodology. (For those not aware of the STOP methodology, essentially it’s every employee’s responsibility to identify and report all areas of safety risk.) Schlumberger realized that in the oil services arena risks are plentiful and thus decided to bring some solid structure behind their process. Their QUEST system manages all global incidents and keeps every Schlumberger employee abreast of the latest incidents and risks, and provides them with an immediate ability to record risks they detect and possible solutions. However, the key component is the escalation… top management is advised of risks which have not been addressed promptly or catastrophic incidents (loss of life, environmental damage, etc).
As I learned more about Schlumberger’s process I also learned that in the oil services industry, Safety (and the environment) are so important that not having solid processes to manage these points can cost contractors their business with big oil (Exxon, Shell, Chevron, etc).
Applying this lesson to Toyota, and all the other Toyotas out there, at the first indication that an accelerator pedal was involved in a loss of life incident, Mr. Toyoda should have been notified. Was he? Probably not, and perhaps it was after several losses that someone, perhaps a data clerk sifting through warranty data noticed there was a possible problem.
In conclusion, with the kind of technologies and applications available on the market, there is no reason why any company should not adopt an idea management software system to streamline their continuous improvement initiatives, but more importantly, keep their workers and customers safe. Companies spend billions of dollars each year on software to keep their inventory and orders under control, why not spend some of that on keeping loss of life and lost time incidents under control? After all, as we learned with Ford and Firestone, and now with Toyota, loss of life can be costly and especially for those that fail to take all measures necessary to protect it.